6th IT Excellence Awards Post Secondary

1
6th IT Excellence Awards(Post Secondary)

• Mobile Application Generator for Route-based
  Location Sensitive Tourist Companion usingChange
  of GSM Cell ID Event

2004/06/09
Lee Kwok Chau, Tommy (leekc1_at_cse.cuhk.edu.hk) Vide
o over InternEt and Wireless Technologies Lab.,
CUHK
2
Agenda

• Traditional Mobile Tourist Directory Solutions
• Location Estimation Using GSM Cell ID
• Tools for Fast Generation of Tourist Companion
  Software
• Sample Applications MTR Travaller, CU Campus Bus

3

• Traditional Mobile Tourist Directory Solutions

4
Location-Based Services (LBS)

• Services are provided based on users current
  location.
• Applicable on different fields
• Navigation (e.g. Driving)
• Shopping / Tourist Guides
• Games / Entertainment,
• Location estimation acts as a mobile tourist
  directory solution where information and
  interaction are triggered by tourists current
  locations.

5
Current Technologies on LBS

• 1) Global Positioning System (GPS)
• Good Accuracy 30-100m
• Poor indoor and urban-area capabilities
• Generally high power consumption
• Expensive hardware at end users
• 2) 3rd Generation GSM (3G)
• 2G-compatible network the majority
• Extra investment in addition to existing 3G
  network
• 3) Modified SIM Card
• Require telco support
• 4) Existing GSM Network
• Fairly advanced
• Expensive and require run-time support from telco

6
GSM Positioning Methods

• TOA (200m - 10km)
• E-OTD (50m - 100m)

• AOA (gtgt150m)
• A-GPS (10m - 50m)

7
Motivation

• Current positioning methods require
• extra cost to existing network / synchronization
  between base stations
• special hardware to end users
• telco-dependent
• Tourist companion software does not require very
  accurate location information
• GSM cell information (e.g. cell ID) is available
  in ordinary GSM handset
• Symbian phone offers programming capability for
  general developers
• Location estimation by GSM cell ID is adopted in
  our project

8

• Location Estimation Using GSM Cell ID

9
Overview of GSM Cell ID Location Estimation

• Each base station has unique location ID and cell
  ID pair
• Main idea each base station can somehow provide
  certain information about a particular location

Location ID 50 Cell ID 2
Location ID 50 Cell ID 3
Location ID 50 Cell ID 4
Location ID 50 Cell ID 1
10
Overview of GSM Cell ID Location Estimation

• Advantages
• Simple implementation passively obtaining
  current registered cell ID
• Applicable on ordinary GSM phones
• Without any support from telco

11
Experimental Results
For Peoples
12
Experimental Results
13
Experimental Results
For SmarTone
14
Experimental Results
15
Conclusion of the Experiment

• Potential difficulties/inaccuracies in 2D Space
• Cell sizes are varying
• Extent of cell overlapping is quite large
• Reliability greatly depends on the deployment and
  configuration of cells by network operators
• A more robust approach is required
• Making use of GSM cell change events
• Applying on 1D path (for route-based LBS)

16
GSM Cell Change Event

• Received signal strength from current registered
  cell is weaker than another, so cell change
  occurs
• Consequences
• More information provided
• More reliable in detecting boundary

17
Using GSM Cell IDs in 1D Space

• A set of multiple cell change events can indicate
  a path

18
MTR Cell ID Data

• Peoples

19
MTR Cell ID Data

• SmarTone

20
MTR Cell ID Data

• Sunday

21
KCR Cell ID Data

• Peoples

22
KCR Cell ID Data

• SmarTone

23
Advantages of New Approach

• Allow to keep track of entering and leaving a
  region
• Become more reliable than purely mapping cell ID
  to location
• Get rid of the problem of inconsistent cell sizes
  and high extent of cell overlapping
• 30-300m accuracy, enough for route-based mobile
  tourist companion software

24
Evolution of Our Positioning Methods
25

• Tools for Fast Generation of Tourist Companion
  Software

26
Application Development Using GSM Cell ID
27
Desire of Fast Tourist LBS Application Generation

• Current issues/overhead in development
• Developers are required to put a great effort in
  the whole development process
• Tourist information is usually the main focus
• Maintaining the software/information update is
  difficult also
• There is a great desire to have a series of
  software tools to generate route-based tourist
  companion software.

28
Definition Reference Point

• Reference point Place taken as a reference on
  target path in the cell data collection process
  (e.g. bus stop, railway stations)

29
Definition Point of Interest

• Point of Interest (POI) Location that is
  interested in the application level (e.g.
  building, tourist spot)

30
Software Development Kit

• It consists of software tools for
• reducing time spent on tourist LBS application
  development
• providing developers with maximum convenience
• Components include
• Cell Snap cell data collection
• Cell Analyzer cell data processing
• Distance Mapper nearest cell searching
• AppGen tourist LBS application generation

31
Architecture of Software Tool Set
32
Cell Data Collection and Processing

• Data collection has to be done manually
• Multiple network operators and routes are
  involved
• Cell data has to be processed and optimized
  before use
• Cell Snap and Cell Analyzer are introduced

33
Cell Snap

• Collection of cell data was done manually in the
  past
• Cell Snap - a tool for automated cell data
  collection
• Different reference points can also be captured
  by phone camera

34
Need of Data Processing

• Photos and data output are separated, so
  developers have to map cell data and reference
  point on their own
• Better presentation of cell data to developers
• Cell data from multiple telcos have to be
  combined together
• A convenient way to merge data for the same route
• Cell Snap data is just raw data
• Data editing tool and automated data analysis

35
Cell Analyzer

• Cell Analyzer provides developers with a
  user-friendly environment to edit, process and
  aggregate cell data into single master data for
  further use

36
Features of Cell Analyzer

• It has the following features
• Data processing
• Data format transform
• Better cell data presentation to developers
• Manual cell data editing
• Data combination for multiple operators
• Automated data optimization
• Cell duplicate removal
• Clear classification between reference points

37
Cell Data in Tree Representation
38
Cell Analyzer in Action
39
Location-to-cell Mapping

• Up to now, pure cell data are not mapped to
  geometrical locations
• It is quite impressive to have nearest target
  searching functions inside tourist companion
  mobile application, such as
• Restaurant
• Toilet
• ATM
• Hospital / Police Station

40
Distance Mapper

• Distance Mapper lets developers to map those POIs
  onto a real map
• The tourist application is then allowed to search
  the nearesttarget from currentlocation
• It reads in the filegenerated byCell Analyzer

41
Application Generation

• The previous tools mentioned enable developers to
  have a nice tourist LBS application development
  environment
• LBS application often focuses also on content and
  service, including actions upon cell changes and
  information provided
• Points of interest (e.g. tourist spots) should
  also be considered instead of just reference
  points (e.g. bus stops)

42
AppGen

• AppGen is specifically designed for content
  builders by generating source code for a tourist
  LBS application
• General developers may also use source code as a
  starting point to build their own applications

43
Options Provided in AppGen

• General
• Map displayed
• Icon
• Message format
• Reference Point
• Location on the map
• Point of Interest
• Name
• Description
• Image
• Path to POI from a reference point

44

• Generated Tourist LBS Applications

45
Generated Mobile Tourist Companion Software

• The mobile tourist application generated would be
  able to
• keep track of interested cell changes and report
  to user (e.g. entering a new MTR station)
• let user specify a POI (e.g. Engineering Building
  in CUHK) and acknowledge user when he/she arrives
  a reference point containing that POI (e.g. bus
  stop)

46
Application in Action (1)
Points of Interest
Reference Points
47
Application in Action (2)
1. User selects adestination
48
Application in Action (3)
2. User is at the startingpoint of a route
49
Application in Action (4)
3. The application keepstrack of the location of
theuser in the path at eachreference point
50
Application in Action (5)
4. Display the informationabout the point of
interestwhen the user arrives thedestination
reference point
51
MTR Traveller and CU Campus Bus

• Make use of GSM cell ID location estimation and
  AppGen in different traffic routes

52
Flow of Tourist Companion LBS Application
Development
53
Summary of Software Tools

• Cell Snap - cell data collection
• Cell Analyzer cell data processing
• Distance Mapper nearest cell searching and
  location-to-cell mapping
• AppGen tourist LBS application generation

54
Advantages of Using LBS Software Tools
Ease location assignment to cell for searching
nearest cell
Generate tourist LBS application
Complexity
Facilitate cell data handling
Work with low-level APIs
Access cell information
Convenience
55
Contribution

• Location estimation can be done by passively
  using GSM cell information
• Applicable on ordinary GSM phones (neither extra
  hardware or special SIM card on user side)
• Telco-independent, stand-alone in any phones
• More robust approach (i.e. cell change event on
  1D path) gives reliable location estimation
• It should be the first complete solution

56
Contribution

• Fast application generation of a route-based
  location sensitive tourist companion software
• Easy to use (hide the concept of GSM cell ID
  location estimation from developers)
• Convenient tools for cell data collection
• User-friendly tools for aggregating cell data
• Multiple network operators (inherited)
• Allow nearest location searching
• Easy content building (nice for tourist apps)

57
Commercialization

• Tourist companion software (with reliable LBS
  available) is highly desired
• University, large company with numerous retail
  shops, public transport (MTR, KCR, tram)
• Nearly everyone are using GSM cell phones
• Smart phones which allow interactive applications
  would become a major trend
• Fast application generation facilitates both
  content providers and software developers

58
Thank you!

• Q A session